Microsoft Certifications in 2026: Eight Current Paths to Consider

Microsoft’s certification portfolio in 2026 looks very different from the catalog many candidates remember from only a few years ago. Several familiar exams retired, AI and security paths changed, and Microsoft continued shifting toward credentials that map to current roles and products. That makes “which Microsoft certification should I take?” a role-selection question rather than a popularity contest.

The Microsoft certification catalog is the right starting point because it lets you compare current exam and certification targets rather than relying on old “top certification” lists. The eight paths below represent different types of work: cloud fundamentals, architecture, DevOps, AI applications and agents, identity, security foundations, data analysis, and database administration.

There is no universal best choice. The strongest certification is the one that validates the work you want to perform next and gives you a practical study plan you can support with labs, projects, or operational evidence.

Microsoft AZ-900: build Azure vocabulary before deeper specialization

Microsoft AZ-900 remains a useful entry point for candidates who need a broad model of cloud computing and Azure. It covers cloud concepts, Azure architecture and services, and management and governance rather than deep implementation.

AZ-900 is most valuable for beginners, business or technical professionals who work around Azure without administering it, and career changers who need the vocabulary used in later role-based certifications. Experienced Azure practitioners may not need it unless they want the credential or need to repair gaps in foundational concepts.

A good AZ-900 project is deliberately small: one resource group, a basic workload, identity and access, storage, monitoring, and governance. The goal is to explain how the pieces fit together rather than become an administrator before taking a fundamentals exam.

Choose AZ-900 when the main gap is understanding Azure as a platform. Do not use it as a substitute for hands-on role training if your target job already requires production administration, engineering, or architecture.

Microsoft AZ-305: architecture for experienced Azure professionals

Microsoft AZ-305 is designed around Azure architecture decisions. Candidates need to reason about identity, governance, monitoring, business continuity, storage, data, compute, applications, and infrastructure while balancing cost, performance, security, and resilience.

This path is much stronger for professionals who already understand Azure administration or engineering than for candidates starting from zero. Architecture questions often contain several technically possible options, and the challenge is choosing the design that best satisfies the stated constraints.

Build preparation around design reviews. Take one workload and produce decisions for identity, networking, compute, storage, observability, recovery, and cost. Then change a requirement such as region, recovery target, compliance boundary, or scale and explain what must change.

Choose AZ-305 when you want to move from operating individual Azure components toward solution architecture and cross-service trade-off reasoning.

Microsoft AZ-400: DevOps engineering across code, pipelines, security, and feedback

Microsoft AZ-400 targets professionals who design and implement DevOps practices using GitHub and Azure DevOps. The current exam places substantial emphasis on build and release pipelines while also covering source control, security, processes, communications, and instrumentation.

AZ-400 is a strong choice when your work connects development and operations: source governance, CI/CD, environments, artifacts, deployment identities, approvals, testing, security controls, monitoring, and feedback loops.

A useful capstone moves one application from work item to source control to tested artifact to protected deployment and then into production telemetry. Break the pipeline deliberately and diagnose which control point failed.

Choose AZ-400 when you want to own software delivery systems rather than only write application code or administer infrastructure independently.

Microsoft AI-103: current Azure AI apps and agents development

Microsoft AI-103 is the current Azure AI Apps and Agents Developer path after the retirement of AI-102 in June 2026. It reflects Microsoft’s current Foundry-centered approach to generative AI applications, agents, retrieval, multimodal solutions, evaluation, safety, and production operations.

This is a better fit for developers and AI engineers than a generic AI-foundation path. Candidates should be comfortable with Python, application architecture, Azure services, identity, data access, and the engineering responsibilities that surround models in production.

Build one application that combines a model, retrieval, an agent or tool, secure identity, evaluation, and monitoring. The important skill is understanding the system around the model rather than only writing prompts.

Choose AI-103 when your target role builds and operates AI applications or agents on Azure and you want a credential aligned to Microsoft’s current AI engineering direction.

Microsoft SC-300: identity and access administration

Microsoft SC-300 focuses on Microsoft Entra identity and access administration. It is a strong path for professionals who work with users, groups, authentication, Conditional Access, workload identities, application access, privileged roles, identity governance, and hybrid identity.

Identity is now a control plane for cloud services, SaaS, applications, and security, so SC-300 skills transfer across Microsoft 365 and Azure environments. The role is especially relevant to identity administrators, security engineers, cloud administrators, and professionals implementing Zero Trust access.

A strong lab follows an identity through its lifecycle: creation, authentication, authorization, access policy, governance review, privilege elevation, and removal. Add a workload identity so the model is not limited to human users.

Choose SC-300 when identity policy and access governance are central to the work you want, rather than treating identity as only one small domain inside a broader security certification.

Microsoft SC-900: security, compliance, and identity fundamentals

Microsoft SC-900 is a fundamentals credential for security, compliance, and identity concepts across Microsoft technologies. It introduces Microsoft Entra, Microsoft security solutions, and Microsoft Purview without requiring specialist implementation depth.

This path suits beginners, business professionals who work with security teams, sales or consulting roles around Microsoft security, and technical candidates who want a structured overview before moving to SC-200, SC-300, SC-401, or another specialist path.

Study SC-900 through business problems: stronger sign-in, least privilege, risky cloud posture, centralized security operations, information protection, data loss prevention, retention, audit, and eDiscovery. Then map each problem to the appropriate Microsoft capability family.

Choose SC-900 when your main need is a coherent security-and-compliance vocabulary. Skip directly to a role-based exam if you already operate those systems and only need deeper implementation skill.

Microsoft PL-300: Power BI data-analysis skills

Microsoft PL-300 targets Power BI data analysts who prepare data, build semantic models, create measures, design reports, and manage and secure Power BI content.

It is a strong certification for analysts, BI professionals, finance or operations specialists, and anyone whose work requires turning business data into governed analytical products. The value comes from the end-to-end workflow rather than visual design alone.

Build one project with imperfect source data, Power Query transformations, a clear dimensional model, DAX measures, a report tied to stakeholder decisions, publication, refresh, workspace governance, and row-level security.

Choose PL-300 when analysis and Power BI are central to your role. It is not a replacement for database administration or data engineering credentials, even though the fields overlap.

Microsoft DP-300: Azure database administration

Microsoft DP-300 validates database-administration skills across Azure and hybrid Microsoft data platforms. The role covers platform planning, security, monitoring, performance, automation, high availability, and disaster recovery.

This path is appropriate for DBAs, cloud database specialists, platform engineers, and professionals who own production relational services. It is much more operational than DP-900 and much more database-focused than PL-300.

A good portfolio project includes platform selection, identity and network security, monitoring, a repeatable performance investigation, an automation task, and a tested recovery workflow. Document the evidence behind every change.

Choose DP-300 when database operations are central to your target role and you want to prove that you can keep data services secure, observable, performant, and recoverable.

Choose the role first, then the certification

These eight paths are not a ladder that every learner should climb in order. They represent different responsibilities. AZ-900 and SC-900 are foundations; AZ-305 is architecture; AZ-400 is DevOps; AI-103 is AI application engineering; SC-300 is identity; PL-300 is analysis; DP-300 is database operations.

Compare your target job descriptions with the current exam objectives. Mark each domain as already strong, needs practice, or new. That produces a clearer plan than selecting whichever certification appears most often in social media or salary lists.

Also compare how you prefer to work. Architects spend more time on trade-offs and design; DevOps engineers on delivery systems; identity professionals on access policy; analysts on data models; DBAs on production data services; AI developers on applications, agents, evaluation, and model-integrated workflows.

Build evidence before you schedule. A small architecture document, pipeline, identity lab, Power BI project, database operations case, or AI application can reveal whether you actually enjoy the responsibilities the credential represents.

Recheck currentness before scheduling because Microsoft changes the portfolio throughout the year. A certification decision should be based on the active credential, current skills outline, and work you want to perform—not a list written for a previous exam generation.

Finally, prefer a coherent story over a large badge collection. Two certifications that clearly support one role, backed by practical projects, usually communicate more value than several unrelated credentials with no evidence of how the skills fit together.

Your roadmap can also change as experience grows. A fundamentals exam that is valuable today may become unnecessary later, while a specialist or architecture credential can make more sense after you have operated the platform in production. Reassess the plan instead of treating the first certification list as permanent.